Differential effects of aging on dendritic spines in visual cortex and prefrontal cortex of the rhesus monkey

Highlights

• Unlike neurons in dlPFC, neurons in V1 do not lose spines with age in rhesus monkeys.

• Aging decreases the proportion of thin spines in macaque dlPFC but not in V1.

• Smaller dlPFC mean spine head diameter correlates with fewer trials to criterion on DR and DNMS tasks.

Abstract

Aging decreases the density of spines and the proportion of thin spines in the non-human primate (NHP) dorsolateral prefrontal cortex (dlPFC). In this study, we used confocal imaging of dye-loaded neurons to expand upon previous results regarding the effects of aging on spine density and morphology in the NHP dlPFC and compared these results to the effects of aging on pyramidal neurons in the primary visual cortex (V1). We confirmed that spine density, and particularly the density of thin spines, decreased with age in the dlPFC of rhesus monkeys. Furthermore, the average head diameter of non-stubby spines in the dlPFC was a better predictor than chronological age of the number of trials required to reach criterion on both the delayed response test of visuospatial working memory and the delayed nonmatching-to-sample test of recognition memory. By contrast, total spine density was lower on neurons in V1 than in dlPFC, and neither total spine density, thin spine density, nor spine size in V1 was affected by aging. Our results highlight the importance and selective vulnerability of dlPFC thin spines for optimal prefrontal-mediated cognitive function. Understanding the nature of the selective vulnerability of dlPFC thin spines as compared to the resilience of thin spines in V1 may be a promising area of research in the quest to prevent or ameliorate age-related cognitive decline.

Introduction

Aging humans and non-human primates (NHPs) often develop cognitive impairment revealed by dorsolateral prefrontal cortex (dlPFC)-dependent cognitive tasks. Two tasks used to evaluate dlPFC function in NHPs are the delayed nonmatching-to-sample recognition memory task (DNMS) and the delayed response test of visuospatial working memory (DR) (Rapp and Amaral, 1989, Luebke et al., 2010). DR performance critically requires dlPFC integrity (Gross and Weiskrantz, 1962, Divac and Warren, 1971), and changes in dlPFC morphological and electrophysiological characteristics correlate with changes in DNMS performance (Peters et al., 1998, Chang et al., 2005, Shamy et al., 2011). Aging rhesus monkeys are generally impaired in both acquisition and performance across increasing memory delays on DR and DNMS (Rapp and Amaral, 1989, Roberts et al., 1997, Rapp et al., 2003, Nagahara et al., 2010), suggesting that dlPFC function may be degraded in aged animals. This can be contrasted with the relative preservation of function in aging humans and NHPs with respect to visual discrimination tasks not reliant on dlPFC (Bartus et al., 1979, Rapp, 1990, Rapp et al., 2003).

The delineation of the molecular and structural alterations that underlie these deficits is an important focus of research in cognitive aging. Normal aging is not associated with significant neuronal loss in the human (Pakkenberg and Gundersen, 1997) or macaque neocortex (Peters et al., 1994). Instead, age-related cognitive decline is thought to result from more subtle synaptic alterations (Morrison and Hof, 1997).

Most excitatory synapses between cortical neurons occur on dendritic protrusions called spines (Nimchinsky et al., 2002). Pyramidal neurons in the macaque dlPFC lose a significant proportion of their dendritic spines with age (Hao et al., 2007, Dumitriu et al., 2010), and morphologically distinct types of spines are differentially affected (Benavides-Piccione et al., 2013). Among spines with a discernable neck (non-stubby spines), the density of spines with large head diameters (mushroom spines) does not decrease with age in rhesus monkey dlPFC (Hao et al., 2007, Dumitriu et al., 2010). Instead, the age-related decrease in spine density is driven by the loss of long, thin spines (Hao et al., 2007, Dumitriu et al., 2010), thought to be highly plastic (Kasai et al., 2010a, Kasai et al., 2010b) and critically important for working memory (Arnsten et al., 2012). We have hypothesized that the “synaptic strategy” underlying function of dlPFC requires the extensive ongoing synaptic plasticity and flexibility that thin spines provide (Morrison and Baxter, 2012). We hypothesize further that while synaptic plasticity also occurs in sensory areas such as the primary visual cortex (Trachtenberg et al., 2002, Gilbert and Li, 2012), the balance between stability and plasticity likely differs from dlPFC, which should be reflected in regional differences in both spine populations and vulnerability to age. Thus, we chose V1 for a regional comparison of age-related effects relevant to synaptic stability vs. plasticity. Other groups have also drawn regional comparisons between V1 and dlPFC in the context of synaptic aging (Peters et al., 1998, Peters et al., 2001, Amatrudo et al., 2012). To address this issue, we imaged pyramidal neurons in layer III of Brodmann areas 46 (dlPFC) and 17 (V1) and examined the density and morphology of dendritic spines along segments of apical and basal dendrites. We found that the density of thin spines is reduced with age in dlPFC, but not in V1, and that the density of other spine types does not change with age in either area.